The goal of the proposed research is to investigate the role of differential motion patterns in the early construction by infants and young children of the concepts animate and inanimate. Understanding and predicting the behavior of members of these categories is crucial to human survival. However, even to begin the construction of such categories, infants must be able to identify entities as potential members of one class or the other. Because animate and inanimate motion are so distinct and because motion provides direct, perceptual evidence of underlying differences between animates and inanimates, the use of motion would provide infants with an excellent starting point from which eventually to construct a more mature understanding of these concepts. Although quite popular, this proposal remains to be directly empirically tested. The first series of proposed experiments employs a habituation of looking paradigm to investigate whether infants have indeed associated distinctive motion patterns with real-world animate and inanimate entities. It further addresses the particular motion characteristics and perceptual properties that most powerfully specify these categories. In the second series, we investigate whether infants represent animates and inanimates as distinct individuals, also using a looking time paradigm. Finally, the third series addresses the development of these abilities over the first years of life by probing the use of motion by young children both to distinguish animates from inanimates and also to infer the presence or absence of specific perceptual, psychological, and biological properties. This research has 3 long-term objectives: 1) To investigate the early developmental course of the construction of the fundamental concepts animate and inanimate. 2) To shed light more generally on the relation between children's sensitivity to perceptual features in the identification of objects as instances of specific categories and their conceptual understanding of these categories. 3) To serve as a methodological pilot and benchmark with normal children with which to investigate the construction of these categories in autistic individuals, for whom the identification of and reasoning about animates might prove particularly problematic.